Parameter tuning method and device, computer equipment, readable storage medium and program product

Abstract

The invention relates to a parameter tuning method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the steps of obtaining historical rule early warning data, determining target parameters needing to be adjusted and optimized according to the historical rule early warning data, and determining core parameters in the target parameters; determining a parameter combination related to a core parameter in the historical rule early warning data; evaluating the trained preset prediction model, and further determining an optimal prediction model; predicting to obtain a value of the service index corresponding to the parameter combination, and determining a corresponding relationship between the parameter combination and the value of the service index; and adjusting the interval of the core parameter in the parameter combination, and taking the interval corresponding to the maximum value of the business index as the optimal interval of the core parameter in the parameter combination. By adopting the method, a high-potential parameter combination which can be obviously influenced can be quickly locked, and perfect balance between tuning efficiency and effect is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data monitoring, in particular to a parameter tuning method and device, computer equipment, computer readable storage medium and computer program product. BACKGROUND

[0002] With the increasingly stringent global financial regulation, anti-money laundering has become the core link of compliance management of financial institutions. At present, financial institutions generally use rule-based monitoring as the first line of defense for anti-money laundering. This method scans the transaction flow in real time by pre-setting a series of static rules (such as "single-day cumulative cash transaction exceeds a certain threshold", "abnormal cross-border remittance frequency in a short period of time", etc.), thereby generating suspicious transaction alerts.

[0003] However, how to effectively integrate traditional rule engines with modern machine learning models and achieve continuous and automated tuning of the entire monitoring system remains a major technical challenge currently faced by the industry.

[0004] In related technologies, the parameters (such as amount threshold, frequency threshold) of existing rule engines are mostly set based on historical experience or regulatory hard rules, lacking data-driven quantitative tuning methods. The tuning process highly depends on manual analysis by business experts, which is not only time-consuming and labor-intensive, but also highly subjective, making it difficult to find a globally optimal solution, resulting in rules that are either too sensitive to produce a large number of false positives or too lenient to increase the risk of missed reports. SUMMARY

[0005] Therefore, it is necessary to provide a parameter tuning method, device, computer equipment, computer readable storage medium and computer program product capable of improving the accuracy of parameter tuning to solve the above technical problems.

[0006] In a first aspect, the present application provides a parameter tuning method, comprising:

[0007] obtaining historical rule warning data, determining target parameters that need to be tuned according to the historical rule warning data, and determining core parameters in the target parameters;

[0008] determining a parameter combination related to the core parameters in the historical rule warning data;

[0009] based on the historical rule warning data and the parameter combination, evaluating a pre-set prediction model after training, and determining an optimal prediction model according to the evaluation result;

[0010] based on the optimal prediction model, predicting the numerical value of the business indicators corresponding to the parameter combination;

[0011] determine a corresponding relationship between the parameter combination and the value of the business index based on the parameter combination and the value of the business index;

[0012] adjust an interval of a core parameter in the parameter combination, and take an interval corresponding to a maximum value of the business index as an optimal interval of the core parameter in the parameter combination.

[0013] In one of the embodiments, the determining of the core parameter in the target parameter comprises:

[0014] adjust the value of the target parameter in a preset interval according to a preset step length, and determine change data of the business index in a case where different target parameters have the same change in value;

[0015] take a target parameter whose change degree of the change data exceeds a preset change threshold as the core parameter.

[0016] In one of the embodiments, the historical rule warning data comprises first historical data in a first business period and second historical data in a second business period; and the training process of the preset prediction model comprises:

[0017] train the preset prediction model based on the first historical data;

[0018] obtain the trained preset prediction model in a case where the accuracy of the prediction result output by the preset prediction model exceeds a preset accuracy threshold and the preset prediction model is verified successfully according to the second historical data.

[0019] In one of the embodiments, the evaluating of the trained preset prediction model and the determining of the optimal prediction model according to the evaluation result comprise:

[0020] calculate a plurality of evaluation indexes of each trained preset prediction model;

[0021] perform weighted summation after assigning a preset weight to each evaluation index to obtain an evaluation result of each trained preset prediction model;

[0022] take a trained preset prediction model corresponding to an evaluation result with a maximum value as the optimal prediction model.

[0023] In one of the embodiments, the evaluation indexes comprise sorting ability and probability calibration degree.

[0024] In one of the embodiments, the method further comprises:

[0025] perform backtracking verification on the optimal interval of the core parameter in the parameter combination;

[0026] If the backtracking verification is successful, the parameter combination will be deployed to the rule engine according to the optimal range.

[0027] Secondly, this application also provides a parameter tuning device, comprising:

[0028] The acquisition module is used to acquire historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters.

[0029] The determination module is used to determine the parameter combinations related to core parameters contained in the historical rule warning data;

[0030] The determining module is further configured to evaluate the trained preset prediction model based on the historical rule warning data and the parameter combination, and determine the optimal prediction model based on the evaluation results.

[0031] The prediction module is used to predict the values ​​of the business indicators corresponding to the parameter combination based on the optimal prediction model.

[0032] The determining module is further configured to determine the correspondence between the parameter combination and the value of the business indicator based on the parameter combination and the value of the business indicator;

[0033] The adjustment module is used to adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

[0034] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0035] Obtain historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters;

[0036] The historical rule-based early warning data is determined to contain parameter combinations related to core parameters;

[0037] Based on the historical rule-based early warning data and the parameter combination, the trained preset prediction model is evaluated, and the optimal prediction model is determined according to the evaluation results.

[0038] Based on the optimal prediction model, the values ​​of the business indicators corresponding to the parameter combination are predicted.

[0039] Based on the parameter combination and the value of the business indicator, determine the correspondence between the parameter combination and the value of the business indicator;

[0040] Adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

[0041] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0042] Obtain historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters;

[0043] The historical rule-based early warning data is determined to contain parameter combinations related to core parameters;

[0044] Based on the historical rule-based early warning data and the parameter combination, the trained preset prediction model is evaluated, and the optimal prediction model is determined according to the evaluation results.

[0045] Based on the optimal prediction model, the values ​​of the business indicators corresponding to the parameter combination are predicted.

[0046] Based on the parameter combination and the value of the business indicator, determine the correspondence between the parameter combination and the value of the business indicator;

[0047] Adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

[0048] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0049] Obtain historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters;

[0050] The historical rule-based early warning data is determined to contain parameter combinations related to core parameters;

[0051] Based on the historical rule-based early warning data and the parameter combination, the trained preset prediction model is evaluated, and the optimal prediction model is determined according to the evaluation results.

[0052] Based on the optimal prediction model, the values ​​of the business indicators corresponding to the parameter combination are predicted.

[0053] Based on the parameter combination and the value of the business indicator, determine the correspondence between the parameter combination and the value of the business indicator;

[0054] Adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

[0055] The aforementioned parameter tuning method, apparatus, computer equipment, computer-readable storage medium, and computer program product first acquire historical rule warning data. Based on this data, they determine the target parameters requiring tuning and identify core parameters within these target parameters. They then identify parameter combinations related to the core parameters within the historical rule warning data. Based on the historical rule warning data and parameter combinations, they evaluate the trained preset prediction model and determine the optimal prediction model based on the evaluation results. Using the optimal prediction model, they predict the values ​​of business indicators corresponding to the parameter combinations. Based on the parameter combinations and the values ​​of the business indicators, they determine the correspondence between the parameter combinations and the business indicator values. Finally, they adjust the range of the core parameters within the parameter combinations, using the range corresponding to the maximum value of the business indicator as the optimal range for the core parameters within the parameter combination. Thus, by determining the core parameters and further confirming the core combinations, they predict business indicators using the prediction model to determine the optimal range, thereby achieving continuous tuning of the rule parameters. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 This is a diagram illustrating the application environment of the parameter tuning method in one embodiment;

[0058] Figure 2 This is a flowchart illustrating a parameter tuning method in one embodiment;

[0059] Figure 3 This is a structural block diagram of the parameter tuning device in one embodiment;

[0060] Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0062] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0063] The parameter tuning method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or located on the cloud or other network servers. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, drones, low-altitude aircraft, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, projection devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. Server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0064] In one exemplary embodiment, such as Figure 2 As shown, a parameter tuning method is provided, which can be applied to... Figure 1 Taking terminal 102 as an example, the explanation includes the following steps 202 to 212. Wherein:

[0065] Step 202: Obtain historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters.

[0066] For example, historical rule warning data is obtained, the target parameters designed in the historical rule warning data are determined and optimized, and the target parameters are adjusted while keeping other target parameters unchanged. Based on the changes in business indicators caused by the adjustment of the target parameters, the core parameters are determined from the target parameters.

[0067] Step 204: Determine the parameter combinations related to core parameters contained in the historical rule-based early warning data.

[0068] Optionally, based on the parameter combinations designed in the historical rule-based early warning data, a parameter combination including core parameters can be determined.

[0069] Step 206: Based on historical rule-based early warning data and parameter combinations, evaluate the pre-trained prediction model and determine the optimal prediction model based on the evaluation results.

[0070] For example, based on historical rule-based early warning data and parameter combinations, the pre-trained prediction model is comprehensively evaluated, and the optimal prediction model is determined based on the evaluation results.

[0071] Step 208: Based on the optimal prediction model, predict the values ​​of the business indicators corresponding to the parameter combinations.

[0072] Optionally, the parameter combination can be input into the optimal prediction model to predict the value of the business indicator corresponding to the parameter combination.

[0073] In one embodiment, the business metric can be the "bad customer rate," which is the ratio of the predicted number of bad customers to the actual number of bad customers in actual implementation. Bad customers refer to customers who may pose a risk of resource misconduct (e.g., money laundering).

[0074] Step 210: Based on the parameter combination and the values ​​of the business indicators, determine the correspondence between the parameter combination and the values ​​of the business indicators.

[0075] For example, based on the parameter combination and the predicted business indicator values, the correspondence between the parameter combination and the business indicator values ​​is determined.

[0076] Step 212: Adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

[0077] Optionally, a Bayesian optimization algorithm is adopted as the core search strategy. Based on the correspondence between parameter combinations and the values ​​of business indicators, the data collection function is used to intelligently suggest the next most likely solution to find the global optimum, that is, the range of core parameters in the parameter combination with the largest data of business indicators, thereby determining the optimal range of core parameters in the parameter combination.

[0078] The above parameter tuning method involves: acquiring historical rule-based early warning data; determining the target parameters to be tuned based on this data, and identifying core parameters within these target parameters; identifying parameter combinations related to the core parameters within the historical rule-based early warning data; evaluating the trained preset prediction model based on the historical rule-based early warning data and parameter combinations, and determining the optimal prediction model based on the evaluation results; predicting the values ​​of business indicators corresponding to the parameter combinations based on the optimal prediction model; determining the correspondence between the parameter combinations and the business indicator values ​​based on the parameter combinations and business indicator values; and adjusting the range of the core parameters in the parameter combinations, with the range corresponding to the maximum value of the business indicator being taken as the optimal range of the core parameters in the parameter combinations. Thus, by determining the core parameters, and further confirming the core combinations, the prediction model is used to predict business indicators to determine the optimal range, thereby achieving continuous tuning of the rule parameters.

[0079] In an exemplary embodiment, determining core parameters among target parameters includes: for each target parameter, adjusting the value of the target parameter within a preset range according to a preset step size, determining the change data of business indicators when different target parameters undergo the same change in value; and taking the target parameter whose change in the change data exceeds a preset change threshold as the core parameter.

[0080] In practice, each target parameter is tested independently in the historical rule-based early warning data. While keeping all other parameters unchanged, the value of the target parameter is adjusted within a preset range according to a preset step size. The changes in the corresponding business indicators of the target parameter are observed. By quantifying the changes (e.g., calculating the rate of change or percentage change), the target parameters whose changes exceed a preset threshold are taken as core parameters.

[0081] In the above embodiments, core parameters mean that even a small change in these parameters will have a significant impact on the business results, and fine sampling is required in subsequent optimization. For other non-core parameters, the impact is smaller, and sparse sampling or even fixing can be performed, thereby greatly reducing the range of parameter adjustment and laying the foundation for efficient optimization.

[0082] In an exemplary embodiment, the historical rule warning data includes first historical data within a first business cycle and second historical data within a second business cycle; the training process of the preset prediction model includes: training the preset prediction model based on the first historical data; and obtaining the trained preset prediction model when the accuracy of the prediction result output by the preset prediction model exceeds a preset accuracy threshold and the preset prediction model is successfully verified based on the second historical data.

[0083] In practice, the preset prediction model is trained using the first historical data; when the accuracy of the prediction results output by the preset prediction model exceeds the preset accuracy threshold, and the preset prediction model is successfully verified based on the second historical data, the trained preset prediction model is obtained.

[0084] The first historical data can be historical rule warning data within the previous business cycle, and the second historical data can be historical rule warning data from the previous business cycle. The preset prediction model is a logistic regression model, a random forest model, and a gradient boosting decision tree model, or it can be other models with prediction functions. This application embodiment does not limit this.

[0085] In the above embodiments, a preset prediction model is trained, and a time-series cross-validation strategy is adopted after training, that is, data from different business cycles are used for validation to simulate the time-series patterns of real business in China.

[0086] In an exemplary embodiment, the trained preset prediction model is evaluated, and the optimal prediction model is determined based on the evaluation results. This includes: calculating multiple evaluation metrics for each trained preset prediction model; assigning preset weights to each evaluation metric and performing weighted summation to obtain the evaluation result of each trained preset prediction model; and selecting the trained preset prediction model corresponding to the evaluation result with the largest value as the optimal prediction model.

[0087] In practice, for each pre-trained prediction model, multiple evaluation metrics are calculated; each evaluation metric is assigned a pre-defined weight and then summed in a weighted manner to obtain the evaluation result of each pre-trained prediction model; the pre-trained prediction model with the largest evaluation result is taken as the optimal prediction model.

[0088] In the above embodiments, through this comprehensive evaluation system, the system automatically selects the best-performing prediction model as the engine for subsequent parameter optimization predictions. This mechanism ensures that the system can adaptively select the prediction tool most suitable for the current data characteristics, improving the adaptability of the solution.

[0089] In one exemplary embodiment, the evaluation metrics include ranking ability and probability calibration degree.

[0090] In practice, the evaluation metrics for the pre-defined prediction model include not only the accuracy of the prediction results (such as RMSE and MAE), but also the ranking ability (such as AUC) and the degree of probability calibration (such as LogLoss).

[0091] In the above embodiments, the model evaluation metrics not only focus on the overall prediction accuracy, but also on the ranking ability and probability calibration degree that are closely related to risk control business, so that the final selected model is more in line with reality.

[0092] In an exemplary embodiment, the parameter tuning method further includes: backtracking and verifying the optimal range of the core parameters in the parameter combination; if the backtracking verification is successful, deploying the parameter combination to the rule engine according to the optimal range.

[0093] In practice, the optimal range of the core parameters in the determined parameter combination is applied to other business cycles for simulation. If it is determined that it will eventually bring the expected performance improvement, then the backtracking verification is considered successful. For the optimal range of the parameter combination that has passed the verification, it is deployed to the rule engine in the production environment for actual decision-making in the next business cycle.

[0094] In the above embodiments, a closed loop is formed by performing a series of analyses, predictions, and verifications on the target parameters, and finally deploying them after verification, enabling the rule engine to continuously, automatically, and securely iterate and optimize according to the business cycle.

[0095] To illustrate the parameter tuning method in this application in detail, an embodiment is described below. For example, this application describes a parameter tuning method in a specific scenario.

[0096] First, obtain historical rule-based early warning data, determine the target parameters designed in the historical rule-based early warning data, and optimize them. While keeping other target parameters unchanged, adjust the target parameters, and determine the core parameters among the target parameters based on the changes in business indicators caused by the adjustment of the target parameters.

[0097] Based on the parameter combinations designed from historical rule-based early warning data, determine the parameter combinations that include core parameters.

[0098] Based on historical rule-based early warning data and parameter combinations, the trained preset prediction model is comprehensively evaluated, and the optimal prediction model is determined according to the evaluation results. The parameter combinations are then input into the optimal prediction model to predict the values ​​of the business indicators corresponding to the parameter combinations.

[0099] Based on the parameter combinations and the predicted business indicator values, determine the correspondence between the parameter combinations and the business indicator values.

[0100] The Bayesian optimization algorithm is adopted as the core search strategy. Based on the correspondence between parameter combinations and business indicator values, the data collection function is used to intelligently suggest the next most likely solution to find the global optimum, which is the range of core parameters in the parameter combination with the largest data of business indicators, thereby determining the optimal range of core parameters in the parameter combination.

[0101] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0102] Based on the same inventive concept, this application also provides a parameter tuning device for implementing the parameter tuning method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more parameter tuning device embodiments provided below can be found in the limitations of the parameter tuning method described above, and will not be repeated here.

[0103] In one exemplary embodiment, such as Figure 3 As shown, a parameter optimization device is provided, including: an acquisition module 301, a determination module 302, a prediction module 303, and an adjustment module 304, wherein:

[0104] The acquisition module is used to acquire historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters.

[0105] The determination module is used to determine the parameter combinations related to core parameters contained in the historical rule warning data;

[0106] The determining module is further configured to evaluate the trained preset prediction model based on the historical rule warning data and the parameter combination, and determine the optimal prediction model based on the evaluation results.

[0107] The prediction module is used to predict the values ​​of the business indicators corresponding to the parameter combination based on the optimal prediction model.

[0108] The determining module is further configured to determine the correspondence between the parameter combination and the value of the business indicator based on the parameter combination and the value of the business indicator;

[0109] The adjustment module is used to adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

[0110] Each module in the aforementioned parameter tuning device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0111] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 4 As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a parameter tuning method.

[0112] The display unit of this computer device is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of this computer device can be a touch layer covering the display screen, or buttons, a trackball, or a touchpad set on the casing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0113] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0114] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0115] Obtain historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters;

[0116] The historical rule-based early warning data is determined to contain parameter combinations related to core parameters;

[0117] Based on the historical rule-based early warning data and the parameter combination, the trained preset prediction model is evaluated, and the optimal prediction model is determined according to the evaluation results.

[0118] Based on the optimal prediction model, the values ​​of the business indicators corresponding to the parameter combination are predicted.

[0119] Based on the parameter combination and the value of the business indicator, determine the correspondence between the parameter combination and the value of the business indicator;

[0120] Adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

[0121] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0122] Obtain historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters;

[0123] The historical rule-based early warning data is determined to contain parameter combinations related to core parameters;

[0124] Based on the historical rule-based early warning data and the parameter combination, the trained preset prediction model is evaluated, and the optimal prediction model is determined according to the evaluation results.

[0125] Based on the optimal prediction model, the values ​​of the business indicators corresponding to the parameter combination are predicted.

[0126] Based on the parameter combination and the value of the business indicator, determine the correspondence between the parameter combination and the value of the business indicator;

[0127] Adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

[0128] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0129] Obtain historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters;

[0130] The historical rule-based early warning data is determined to contain parameter combinations related to core parameters;

[0131] Based on the historical rule-based early warning data and the parameter combination, the trained preset prediction model is evaluated, and the optimal prediction model is determined according to the evaluation results.

[0132] Based on the optimal prediction model, the values ​​of the business indicators corresponding to the parameter combination are predicted.

[0133] Based on the parameter combination and the value of the business indicator, determine the correspondence between the parameter combination and the value of the business indicator;

[0134] Adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

[0135] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0136] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0137] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0138] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A parameter tuning method, characterized in that, The method includes: Obtain historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters; The historical rule-based early warning data is determined to contain parameter combinations related to core parameters; Based on the historical rule-based early warning data and the parameter combination, the trained preset prediction model is evaluated, and the optimal prediction model is determined according to the evaluation results. Based on the optimal prediction model, the values ​​of the business indicators corresponding to the parameter combination are predicted. Based on the parameter combination and the value of the business indicator, determine the correspondence between the parameter combination and the value of the business indicator; Adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

2. The method according to claim 1, characterized in that, The process of determining the core parameters from the target parameters includes: For each target parameter, the value of the target parameter is adjusted within a preset range according to a preset step size to determine the change data of business indicators when different target parameters change in the same way. The target parameter whose degree of change in the changed data exceeds a preset change threshold is used as the core parameter.

3. The method according to claim 1, characterized in that, The historical rule-based early warning data includes first historical data within the first business cycle and second historical data within the second business cycle; the training process of the preset prediction model includes: The preset prediction model is trained based on the first historical data; If the accuracy of the prediction result output by the preset prediction model exceeds the preset accuracy threshold, and the preset prediction model is successfully verified based on the second historical data, the trained preset prediction model is obtained.

4. The method according to claim 1, characterized in that, The step of evaluating the pre-trained prediction model and determining the optimal prediction model based on the evaluation results includes: For each pre-trained prediction model, calculate multiple evaluation metrics for each pre-trained prediction model. After assigning a preset weight to each evaluation indicator, a weighted sum is performed to obtain the evaluation result of each trained preset prediction model. The preset prediction model trained with the largest evaluation result is taken as the optimal prediction model.

5. The method according to claim 4, characterized in that, The evaluation metrics include ranking ability and probability calibration degree.

6. The method according to claim 1, characterized in that, The method further includes: The optimal range of the core parameters in the parameter combination is backtracked for verification; If the backtracking verification is successful, the parameter combination will be deployed to the rule engine according to the optimal range.

7. A parameter tuning device, characterized in that, The device includes: The acquisition module is used to acquire historical rule warning data, determine the target parameters that need to be optimized based on the historical rule warning data, and determine the core parameters among the target parameters. The determination module is used to determine the parameter combinations related to core parameters contained in the historical rule warning data; The determining module is further configured to evaluate the trained preset prediction model based on the historical rule warning data and the parameter combination, and determine the optimal prediction model based on the evaluation results. The prediction module is used to predict the values ​​of the business indicators corresponding to the parameter combination based on the optimal prediction model. The determining module is further configured to determine the correspondence between the parameter combination and the value of the business indicator based on the parameter combination and the value of the business indicator; The adjustment module is used to adjust the range of the core parameters in the parameter combination, and take the range corresponding to the maximum value of the business indicator as the optimal range of the core parameters in the parameter combination.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images